Prospective review of SPEP orders confirms need for testing algorithm.

Abstract

e19326 Background: Serum protein electrophoresis (SPEP) identifies monoclonal proteins associated with plasma cell disorders including multiple myeloma. There are no large prospective studies looking at clinical ordering patterns of SPEPs. This study will determine the percent of monoclonal proteins detected by SPEP testing and factors that are or are not associated with identifying monoclonal proteins. We predict that SPEP testing is commonly ordered and not thoroughly evaluated prompting a need for an algorithm to ensure appropriate use and follow up testing. Methods: In our academic center, we captured data on all SPEP tests ordered from March 17 to June 16, 2019. We chart reviewed each associated patient to collect the following: patient age and sex, SPEP result, hemoglobin, serum albumin, total serum protein, serum calcium, serum creatinine, presence of proteinuria, serum free light chains, immunoglobulins (IgG, IgA, IgM), serum immunofixation, urine protein electrophoresis, urine immunofixation, bone marrow biopsy report, imaging results (bone survey, PET, MRI) and echocardiograms. We ascertained the reason for ordering the SPEP, inpatient or outpatient status of the patient and the ordering department. Results: There were 929 SPEPs ordered. 410 (44%) ordered on “new” patients with no previously identified M-spike. Of the 410 “new” SPEP orders, 57 (14%) detected a M-spike. 8 (2%) led to a diagnosis requiring treatment: 7 new cases of myeloma and 1 case of AL amyloidosis. 108 (26%) of the 410 “new” SPEP orders were ordered in the inpatient setting with 20 “new” M-spikes identified. Only 4 of the 108 inpatient SPEP orders identified a diagnosis requiring treatment—4 cases of myeloma. These patients had SPEPs ordered due to pathologic fractures related to lytic lesions and acute renal failure with other associated CRAB features. Neurologic findings were the most common reason for ordering “new” SPEPs, 106 (26%) of the 410 with only 7 with M-spike—all low risk MGUS. Only 2 of 26 patients were found to have low risk MGUS when SPEP was ordered as part of a rheumatologic work up. All 6 patients with SPEP ordered for dermatologic reasons were negative for M-spike. 23 (40%) of the 57 “new” M-spikes had no follow up testing to determine the clinical significance. Conclusions: This prospective review on the use of SPEP orders in an academic medical center documents the need to provide an algorithm to aide effective use of testing to decrease inpatient ordering and provide guidance on appropriate follow up testing. We found a very low yield with SPEP testing for neurologic, rheumatologic and dermatologic reasons.